20 Comments
User's avatar
Kevin M.'s avatar

I don't understand how you could write an article about "who wins Nobel prizes" and not at least mention the fact that over 20% of Nobel prizes have been awarded to Jewish people. That statistic is more remarkable than anything else you mentioned.

EBS's avatar

Not an unrelated fact to most of the German prizes being for work done before 1935

Boyuan Xiao's avatar

Been interested in this phenomenon myself

The strongest case against it is that a lot of Jewish Nobel laureates are only partially Jewish, if you were to look for (say) Scottish or Irish heritage for Prize winners I suspect you'll also find numbers disproportionate to the overall population of those groups

Kaleberg's avatar

What do you mean by partially Jewish? Intermarriage with Ashkenazic Jews was rare until relatively recently, so odds are both parents were Jewish, though, in Jewish law, only the religion of the mother matters. I doubt intermarriage had a big effect.

There's also the issue of conversion. For example, Fritz Haber who won in 1918 had converted to Christianity but had been born Jewish, not that his conversion helped him when the Nazis took over. In a lot of Europe, Judaism is more a matter of race than persuasion.

Scotts are over-represented in engineering. Jews are over-represented in science.

Kevin M.'s avatar

I doubt that's the explanation. IQ test constistently show Jewish people average about 15 points higher than the general population.

Kelly Papapavlou's avatar

What do you mean of Jewish? The religion or the place of origin? Then what about the remaining 80% ? Perhaps this 80% is Christians, Muslims or agnostic......

Mitch R.'s avatar

Then show that on a per-capita basis!

Paul Novosad's avatar

You might be interested in our paper on the childhood socioeconomic status of the Nobel laureates in the sciences, which we proxied based on fathers' occupation: https://paulnovosad.com/pdf/nobel-prizes.pdf

Kaleberg's avatar

That's a good counter-argument against those who claim we need a Trillion Americans to get the brain power needed to solve our nation's problems.

Michael Frank Martin's avatar

This is really important work, and I really appreciate your publishing the dataset. I did a spot check of one prize I have some personal experience with, which is W.E. Moerner's prize for single molecule spectroscopy, which he received in 2014. The dataset lists his affiliation at the time of doing the work as Stanford University. In fact, Moerner had started the work and published many of its key results while at IBM Research in the 1990s and at UCSD, where he moved from IBM before moving to Stanford.

It's just one data point, but it does make me question the integrity of the conclusions that can be reached on the set as a whole. This is important research that deserves more careful scrunity.

Daniel Williams's avatar

Surprising that Russia is not on the countries list as it has spawned so many brilliant scientists. But many of them emigrated out of there.

Ian Keay's avatar

I was thinking the same thing. Could be the anglophone bias, could be that authoritarian regimes stifle original research. Be interesting to update this analysis in 25 years time and see if the US has fallen off a cliff. The UK is building the East West Rail between Oxford and Cambridge, https://en.wikipedia.org/wiki/East_West_Rail, and is seriously considering the Oxford-Cambridge Growth Corridor, https://en.wikipedia.org/wiki/Oxford–Cambridge_Arc, which could if built (big IF) lead to Oxbridge maintaining its position.

Ryan Davidson's avatar

Two quick observations about the universities getting lots of Nobels:

1. Cambridge has been a leading center for research in mathematics and physics since at least the seventeenth century. Both Isaac Newton and Stephen Hawking were at Cambridge, for instance. Oxford has historically been a leader on the humanities side of things, particularly language, literature, and history. As Nobel Prizes aren't awarded in the humanities, it makes sense that Oxford would lead Cambridge here.

2. The reason CalTech and U.C. Berkeley have so many Nobel Prizes is because both are major centers of elementary/particle physics funded by the U.S. federal government. Several major federal atomic research labs (Lawrence Livermore, Sandia, JPL etc.) are closely associated with them, and many leading researchers in things like the Manhattan Project worked at one or the other (e.g., Oppenheimer).

Gary Mindlin Miguel's avatar

> As Nobel Prizes aren't awarded in the humanities

The analysis excluded the nobel in literature, but it exists!

Destiny's avatar

I would have liked to see literature and economics included in your analysis. It would have made the spread less asymmetric. You don’t need a hadron collider to write a story and Oxford is strong in literature. The Peace prize would make it even less asymmetric. Math too but you would need fields medal or Abel prize winners.

firehat's avatar

The absence of Russian/Soviet awards is suspect at best. If the numbers are taken as an indicator of national capacity for innovation or ingenuity they have to be considered suspect.

Kaleberg's avatar

There's a big gap between 1905 and 1956 with only one Russian winning the Nobel Prize in literature in 1933. There was a lot of political instability in Russia during that period. Stalin died in 1953, so things opened up a bit after that, but otherwise Russia was cut off from the world scientific network.

Kaleberg's avatar

For a further example of institutional concentration consider that Rosalyn Yalow and Gertrude Elion had the same chemistry teacher at Walton High School , Mr. Mondzak, before they were admitted to Hunter College.

Agustín's avatar

Hi. Argentina has 4 (four) Nobel prizes. You noted only 2 but it should be 3 in medicine, and chemistry: Bernardo Houssay (Medicine, 1947), Luis Federico Leloir (Chemistry, 1970), and César Milstein (Medicine, 1984).

Octavi Semonin's avatar

Some academics also think the longer delay to giving the award is to avoid PR problems like James Watson. Not saying it’s a good reason, but it might tend to distort the data a bit.